Abstract :
The problem of threshold or weak-signal detection in highly nonGaussian EMI is extended to vector fields, and narrow-band signals and interference. The emphasis is on a canonical formulation, illustrated by a number of specific examples. Spatial sampling with adaptive beam forming, as well as temporal sampling and all relevant vector field components, must be included for maximum processing gain. New results for a canonical theory of these vector detection cases are presented. Jointly and asymptotically locally optimum algorithms and performance measures are obtained. These results provide statistical-physical models of the EMI environment, and they include first-order probability distributions of vector EMI noise fields and received processes, with specific examples of EMI fields generated by randomly distributed electric and magnetic dipole sources, as well as more general sources. The effects of beamforming, selfdirecting beams, multiple field components, fading, and Doppler `smear´ on signal detectability are included
Keywords :
electromagnetic compatibility; electromagnetic interference; information theory; signal detection; Doppler smear; EMC; adaptive beam forming; algorithms; canonical formulation; canonical theory; fading; first-order probability distributions; magnetic dipole sources; multiple field components; narrow-band signals; nonGaussian EMI; performance measures; processing gain; selfdirecting beams; signal detection; statistical communication theory; statistical-physical models; temporal sampling; threshold detection; threshold vector field detectors; vector EMI noise fields; vector field components; weak-signal detection; Detectors; Distributed power generation; Electromagnetic interference; Magnetic field measurement; Magnetic noise; Narrowband; Noise generators; Probability distribution; Signal sampling; Working environment noise;
